spec-planner
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill is subject to potential indirect prompt injection through its discovery mechanism.
- Ingestion points: The
exploresubagent in Phase 2 (DISCOVER) reads local files, modules, and existing patterns to provide context for the specification (SKILL.md). - Boundary markers: There are no explicit instructions to use delimiters or ignore embedded instructions when the agent or its subagents process the content of these files.
- Capability inventory: The skill possesses the ability to spawn subagents using the
Tasktool and write new markdown files to thespecs/directory (SKILL.md). - Sanitization: No sanitization or validation logic is defined for the data retrieved during the exploration phase before it is used to generate the final spec.
- [NO_CODE]: The skill consists exclusively of Markdown-based instructions and reference documentation. No executable scripts (e.g., Python, JavaScript) or binaries are included in the package.
Audit Metadata